shaniaki/linescan-cuda
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
Repository files navigation
The parallel version of the drop detector part of the linescan application by CUDA.
src/ the source directory
drop_detector.cu the entry point and main file of the application
drop_detector_cpu.cpp the reference serial implementation
drop_detector_v1.cu each nozzle associated with a CUDA block
drop_detector_v2.cu each nozzle associated with a CUDA block and CUDA
threads parallelize operations within each blocks
drop_detector_v3.cu parallel version of v2 plus use of shared memory
when possible
inputs/ directory for the input files
each input data set contains 4 files
file_io.cpp functions for reading the inputs from the files
globals.hpp global definitions
results.txt contains examples of execution results
README this file
Notes:
- For each data set, the macro definitions N and BUFFER_SIZE should be adjusted
in the globals.hpp file accordingly.